The McCollough effect (ME) has been shown to be sensitive to cholinerg
ic agents, being strengthened by hyoscine (antagonist) and weakened by
physostigmine (agonist), and possibly to more generalized changes in
CNS arousal. We therefore expected the ME to be sensitive to hormonal
changes during the menstrual cycle, being strongest in the postovulato
ry phases when arousal is low. In two experiments we found a highly si
gnificant effect of menstrual phase for the normally cycling women, bu
t not for oral contraceptive users: ME strength gradually increased ac
ross the cycle, reaching a premenstrual peak. These findings may be ex
plained in terms of hormonally mediated changes in arousal across the
menstrual cycle.